Skip to main content
FOCIS 2025 Main banner
Final Program


Final Program

Autoimmune Diseases

Tu111 - What Can Urine Single-cell Transcriptomics Teach Us About Kidney Immune Responses in Lupus Nephritis?

Tuesday, June 24, 2025
6:00pm - 7:30pm East Coast USA Time
Abstract Text: Immune cells are found in urine samples of lupus nephritis (LN) patients with active disease. We have previously showed, in a small-scale study, that gene expression in urine leukocytes in LN is highly similar to that in their kidney counterparts. Here, we generate single-cell RNA-sequencing (scRNA-seq) data from 140 LN patients, profiling both kidney and urine samples. We show that most leukocyte subsets found in LN kidneys can also be identified in urine. Using spatial transcriptomics data, we show that macrophage subsets that have a preference to localize to glomeruli are overrepresented in urine, suggesting that the glomeruli are a main gateway to urine for leukocytes. While the frequencies of leukocyte subsets in urine do not directly reflect those in kidney, we show that with simple linear modeling we can estimate kidney frequencies from urine frequencies with a median relative error of ~30% for most myeloid subsets. Furthermore, we show that the frequencies of specific urine leukocytes significantly correlate with the NIH Activity and Chronicity indices and the ISN class. Using linear regression models, we can estimate the Activity, Chronicity and ISN class from the frequencies of urine leukocytes with an AUC of 0.8-0.85. In addition, using this approach we can predict a 1-year treatment outcome with an AUC of 0.74. For comparison, models using the urine protein-to-creatine ratio (UPCR) yield an AUC of 0.54-0.68. Our work delineates the usefulness of urine profiling in monitoring kidney immune processes in LN nonintrusively, and provides a valuable resource to LN researchers.